How We Doubled Our Android Install Rate in One Hour | Hacker News Reader

How We Doubled Our Android Install Rate in One Hour

dwellable.com

105 pointsgurgeous23 comments

It helps that the design of the new Dwellable is very similar to the popular magazine Dwell. I wonder how much of the gain is due to brand recognition of a different brand?

- Similar san-serif lowercase font

- Large-format images of modern homes

- Almost-identical name :-)

Of course, you should always be A/B testing, and congrats on the conversion increase.

[edit: formatting]

If one data point matters: I occasionally rent vacation homes and am probably in the target demographic for that app. But I've frankly never heard of "Dwell" and certainly wouldn't recognize their brand artwork.

This is an app targetted at essentially the whole of the US upper middle class; is copying a niche magazine really going to matter much?

> This is an app targetted at essentially the whole of the US upper middle class;

That is pretty much Dwell's market demographic, too.

Sure, but that's not the same thing as saying everyone in the demographic has seen it. Wikipedia has an old number giving their circulation at 260k. That's just noise. There's no benefit gained by copying their artwork, so I think positing that as the reason for the effect just has to be wrong.

Now, this would change if this was an app for a small niche market and the magazine targetted the same people. Then the target users would likely have seen it and there would be benefit. But not here.

A/B testing on Google Play is fairly difficult, as you can only ever have one version uploaded and it takes a few hours for any changes to be made available publicly.

i've never seen or even heard of dwell, but my eye would have glided over the first image without even noticing it, and been arrested by the second one.

I'm not sure if it is the artwork or the message. The previous promo art had tagline "Vacation rentals and reviews" which targets a specific audience, while second artwork just says "Dwellable" which is more mysterious and so you might have more people download it to check it out. Maybe it helps to have a broader message to get more people to try it, but expect to have a higher % of those curious windoshoppers try it and then shelve it/uninstall it.

That's the way it seems to me as well. With the second one that has no tag line I'd think that Dwellable might be a competitor for Trulia, Zillow, or Redfin.

Note that the most important factor here is that they already were being featured on the front page. Making awesome promo art is a good idea, but if they weren't featured on the front page, this promo image wouldn't make a whole lot of difference.

I disagree: it seems clear that the CTR on the new design was 2x the old. Now, granted that 2x of a featured spot is a much bigger absolute number than 2x of typical placement. But 2x is 2x. This is a great reminder of the impact of design for anyone trying to optimize their app store presence (large or small).

I'm not sure, honestly: I imagine the differential of a new design varies depending on how users come across it. If when not being featured Dwellable mostly gets new installs from word-of-mouth, say, I bet that the new design doesn't have as much impact. (I'm sure it would still be a big difference, but maybe on the order of 1.5x instead of 2x, say.)

As they are being featured their promo art is shown to users before they get to the page for that particular app (see the screenshots for their new and old images - you can see it is part of a scrolling bar of promo images).

If they weren't being featured, the promo art would only be seen by users that clicked through to the app's page - users would only have seen the app's icon before getting the app's landing page.

So I think featured vs not featured makes a big difference to the importance of the promo art.

>"instantly DOUBLED our install rate"

They are talking about their install rate not CTR.

This just shows how pictures communicate more efficiently than text. On a busy screen filled with apps, the words "Vacation reviews, photos and more" don't get read, they just get tuned out. But a picture of a beautiful vacation home grabs the eye. There's almost an automatic "I want that" response.

Without telling us the install rate, "doubled" is just lying with statistics. Did you go from 10 to 20 daily or from 10k to 20k? Play says that you have 10k+ installs total (not a lot), so it looks more like the former.

If your initial number is really low, minimal work can double it. If you are doing well, the same work may give you a small percentage improvement, or even nothing. It's called "law of diminishing returns."

What I learn from this post: you had lots of low-hanging fruit, and probably still do.

Really simple and effective example, thank you for sharing!

I wonder how much of it has to do with the objective "quality" of the design versus the effect on the audience's psychology of connecting "real life" (vacation rentals) to an app that lives in your phone.

I'd love to see further experimentation with:

1. Lower quality "real life" asset designs.

2. Higher quality abstract designs.

I'm sure that's a huge part of it. Aspirational vs. functional. Since the app was designed to be in large part aspirational, the new artwork actually matches our intent better than the orange monstrosity.

I don't think I learned anything from the blog post. It was already being featured by Google. May be this is just an ad.

Inept designer here. Happy to chat.

Inept designer here too. It's really awesome to see a success story like this, it gives me hope

Way to advertise yourself :)

I'm an engineer/entrepreneur type guy. Everything else is just by accident.

I don't buy the inept designer act. Your project page is very good and your new promo artwork is like a miniature version of it: http://www.dwellable.com/h/26684/Oahu/East-Oahu/Kailua/Alii-...

Reply on news.ycombinator.com